404 ON THE GENUS CORDULEPHYA, 



we may suppose that the members of the genus were all 

 double-brooded. Both the coastal and mountain-forms then, 

 though they may not at that time have been specifically dis- 

 tinct, were faced with their own most formidable foes in the 

 shape of numerous ravenous Aesrhii'ul larvcC. On the coast, 

 the swarms of larvae of Aeschna hrevistyla gradually exter- 

 minated the early brood of C. jyjlfimiea, while, on the moun- 

 tains, the late-developing larvae of A usfroaesr/uia jicrvistiymn 

 destroyed the second brood of C. montand. That this one 

 circumstance, in itself, may have played a large part in the 

 differentiation of the two species is very probable. 



I find, on referring to my notes on ('. imgmaa, that this 

 theory is supported by the fact that, in October, 1907, I 

 dredged, from the creek at Lily Vale, several very small 

 larvae of ('. jjyfimmt, which were, however, fully developed. 

 I did not know at the time to what species they belonged, 

 but the fact that their wing-cases were so long, reaching 

 nearly to the end of the abdomen, struck me as being so 

 peculiar, that I made a note of it. On examining the 

 labium, I determined the larvae as those of C. 'pygmeea. Now 

 I have never seen the imago out before February. It is not 

 unreasonable, therefore, to suppose that many of the larvae 

 produced from the late brood, feed up rapidly, and are in a 

 fair way to become a first brood for the next season. But 

 here the enemy, the rapacious Aeschna larvae, has to be 

 reckoned with. Those that escape him, must hide away in 

 obscure corners, and probably have to undergo a prolonged 

 fast for many weeks, vmtil the Aeschna larva has become 

 full-fed and emerged. Supposing that the ecdyses take place 

 as usual, we should then have the spectacle of a poor, miser- 

 ably undersized larva, with huge wing-cases, — in fact, just 

 such a one as those I took at Lily Vale. These larvae most 

 certainly did not emerge until the end of February, for I 

 visited Lily Vale every month up to April, and saw no 

 imagines until the beginning of March. Here also may be 



